Bouting inside clubs for optimum increase of fencing ability

[brutus]

all this seems to be a very complicated way of planning training sessions.
Is it not enough to just say that a fencer should be thinking of fencing say 2 better fencers and 2 bunnies, or 3 better fencers and 2 bunnies, etc per session?

Also, I don't like the idea of fencing regulation bout lengths in my practice. I like to train for longer, somtimes 5 or ten mins longer, this enables me to have 'a little bit eft in the tank' incase things go paershaped during competition.

[sabreteur]

Jeez Peter, just require fencers to wear remote controlled shock collars so coaches can 'correct' bad fencing the moment it happens.

I'd prefer that to the brain damage of trying to follow an elaborate fencing efficiency ritual.

[veeco]

all this seems to be a very complicated way of planning training sessions.
Is it not enough to just say that a fencer should be thinking of fencing say 2 better fencers and 2 bunnies, or 3 better fencers and 2 bunnies, etc per session?

Also, I don't like the idea of fencing regulation bout lengths in my practice. I like to train for longer, somtimes 5 or ten mins longer, this enables me to have 'a little bit eft in the tank' incase things go paershaped during competition.

I am not sure this is a good assumption.

If you don't fence to 5 or 15, in your training sessions, and "fence for time" some people say that it actually doesn't amount to a good practice, because you're not trying hard enough. In other words, even if you're willing to accept losing a bout during practice because you're working on some technique or something, you will not be trying hard enough to win if you're not keeping track of the score, and only fencing for time.

[PeterGustafsson]

Hi!

Jeez Peter, just require fencers to wear remote controlled shock collars so coaches can 'correct' bad fencing the moment it happens.

I'd prefer that to the brain damage of trying to follow an elaborate fencing efficiency ritual.

What the heck???

The fencers do not have to think about this, it can be done by the coach alone - even before the training starts, if he knows who will show up for training. No need to tire your appearently easily exhausted head with any *why* thinking, just do as he says and fence the one listed against you. How hard can it be?

OTOH, if you do not like to think, maybe some other sport like powerlifting or the 100m dash is more for you.

The problem I presented in the threadstart is an optimization under constraints problem, a problem class which appears in many endeavours. Both engineering and economics are full of them, for example. The huge practical importance of the problem class and the fact that many of the problems are such that it can not be proved, or is very hard to prove, that the solution which presently seems to be best actually is the best makes the problem class intellectually challenging, and it is satisfying to find a (preferably simple) algorithm which generates solutions which always are close to the true optimal solution.

Or atleast it is to me. Pray tell, what do intellectually challenging problems do you find satisfying to solve? During the waking time when you do not have to use your brain to deal with stuff at hand, do you not try to solve whatever interesting and taxing problems that life throws your way?


Have a nice time!

Peter Gustafsson

[PeterGustafsson]

Hi!

Your point above reminds me that the Team Format is now different. A long time ago the team match was just a series of independent 5-point individual bouts. I was not thinking about the "new format" when I wrote this.

I have also fenced team bouts under that old system. I think the newer is better. For instance, a fencer who is clearly weaker than all of the fencers on his opposing team could not do anything for his team under the old system. Under the new system, he can do his best not to lose by too large a margin. No matter what the fencing strengths are in a given team bout, both fencers can affect the outcome substantially by either fencing the best of their normal ability, or the worst. No so under the old system.

As a practical solution I would still suggest dropping the constraint to finish in one meeting, however if I now view the problem you propose as a more academic excercise and remain within the constraints:

In one session
2 pools of 3 fencers

pool 1 ABC 3 bouts
pool 2 DEU 3 bouts

In the next session the best fencer of pool 2 and worst fencer of pool 1 exchange positions.

Great! We now have a quite good solution to the special problem that I posed the threadstart. In the special problem, there are 15 possible bouts of which 6 should be chosen. There are 5005 different solutions to this, not counting bout order of those bouts which are selected. (If one does count bout order, one gets a little over 3.6 million distinct solutions)

Obviously, which that many possible solutions to a fairly small problem, going through all possible solutions to find the best becomes completely infeasible. Still, finding a good, if not the best, solution - in a quick way - to the general problem would be useful.

Let´s hear the suggestions!


Have a nice time!

Peter Gustafsson

[brutus]

I am not sure this is a good assumption.

If you don't fence to 5 or 15, in your training sessions, and "fence for time" some people say that it actually doesn't amount to a good practice, because you're not trying hard enough. In other words, even if you're willing to accept losing a bout during practice because you're working on some technique or something, you will not be trying hard enough to win if you're not keeping track of the score, and only fencing for time.

yeah sorry, I missed a bit out, I meant to say I don't always fence for regulation length or hits, somtimes I fence for time, mainly to build fight endurance and also to eleviate the pressure of havign to meet a set score, thus enabling a bit more freedom to try different things.

[Durando]

yeah sorry, I missed a bit out, I meant to say I don't always fence for regulation length or hits, somtimes I fence for time, mainly to build fight endurance and also to eleviate the pressure of havign to meet a set score, thus enabling a bit more freedom to try different things.

I suspect you know this already, but it probably bears repeating that time is an (or the) essential dimension of tactical problems. You can't productively practice tactics if you aren't on the clock. "Trying different things" isn't the same as "trying to win the bout."

If you want to build endurance, fence more three minute bouts. Or fence a sequence like best of 1-3-5-10-15-10-5-3-1 touch bouts at 3 minutes for the 1, 3, and 5-touches, six minutes for the 10-touch, and nine minutes for the 15-touch.

[brutus]

I suspect you know this already, but it probably bears repeating that time is an (or the) essential dimension of tactical problems. You can't productively practice tactics if you aren't on the clock. "Trying different things" isn't the same as "trying to win the bout."

This all depends on the tactics you are "trying out". Also you can try out moves, and various actions that wouldn't be be dependant on being in an "on the clock" situation.

[Durando]

This all depends on the tactics you are "trying out". Also you can try out moves, and various actions that wouldn't be be dependant on being in an "on the clock" situation.

You are practicing technique then, not tactics.

[brutus]

You are practicing technique then, not tactics.

Not always, there are short term tactical decisions which will play out in a short period of time as well as long term decisions that will play out over the course of the fight.

Fencing for time allows you to try and experiment with various different tactical situations.

[lemon__fresh]

assuming you can come up with a "valid" algorithm or one able to be executed by a computer(see turing machine) you could write a simple program to come up with the pairings for the bouts

therefor sparing your brain the challenge (although comming up with the program could pose a challenge in itself)

but im sure it could be done without to much hairpulling/keyboard smashing in python, ruby or some other highlevel language

not sure what the actual "ring-with-A/U" algorithm is, but the name sounds familiar

[sabreteur]

Pray tell, what do intellectually challenging problems do you find satisfying to solve?

I'm working on a magic potion that will make coaches from competing clubs and division officials all treat each other with respect and kindness. And flying ponies.



PS. Sorry for the harsh tone of the post. I was aiming for humor and missed. My best guess on what works best (at least for sabre) is three people per strip. Ten touch bouts. Each fencer is up for two bouts, down for one. Every three cycles -or so- one of the fencers switches to another strip.

The bouts between the elites and the bunnies don't take up too much time, and you get so much strip time exhaustion is the main limiter. Anything more than five fencers per strip and you get way too little strip time to get much practice. So the secret to most efficient training is to minimize the fencer/strip ratio.

BTW, anybody care to chime in on what format for practice their club uses?

[PeterGustafsson]

Hi!

My best guess on what works best (at least for sabre) is three people per strip. Ten touch bouts. Each fencer is up for two bouts, down for one. Every three cycles -or so- one of the fencers switches to another strip.

The bouts between the elites and the bunnies don't take up too much time, and you get so much strip time exhaustion is the main limiter. Anything more than five fencers per strip and you get way too little strip time to get much practice. So the secret to most efficient training is to minimize the fencer/strip ratio.

If you refer back to the threadstart, you will notice that a defining characteristic of the scenario that I put forth was that strip time was at a premium, which translates to a high fencer/strip ratio.

In other words, you are saying that the solution is to change the scenario, rather than to work within it. Fine, if one has the resources to do so. Less so if one does not (at least in the short term). Also, solving a problem by dumping resources on it reduces the amount of resources available to other problems in the future and takes away the mental challenge of the problem.


Have a nice time!

Peter Gustafsson

[the ancient one]

"... if I now view the problem you propose as a more academic excercise and remain within the constraints:

In one session
2 pools of 3 fencers

pool 1 ABC 3 bouts
pool 2 DEU 3 bouts

In the next session the best fencer of pool 2 and worst fencer of pool 1 exchange positions.

I (ab)used the gaussian curve to model an accumulation of experience. An xls file is attached.

Using the above scheme, but WITHOUT switching the low fencer of the high group with the high fencer of the low group each session, I model an increase of experience over several meetings. As you could anticipate the highest fencer gains the least and the lower fencer gains the most. The constants of the curve can be played with. Perhaps later if work is really boring me I will make the full scenario where I follow my rule for rotating fencers between groups and compare the growth rates.

I thought about a vector approach also, but how many dimensions to use:
for example (a,b) where a is an offensive ability and b is a defensive ability. Tactics? Speed? so many possibilities.

[oso97]

PS. Sorry for the harsh tone of the post. I was aiming for humor and missed. My best guess on what works best (at least for sabre) is three people per strip. Ten touch bouts. Each fencer is up for two bouts, down for one. Every three cycles -or so- one of the fencers switches to another strip.

The bouts between the elites and the bunnies don't take up too much time, and you get so much strip time exhaustion is the main limiter. Anything more than five fencers per strip and you get way too little strip time to get much practice. So the secret to most efficient training is to minimize the fencer/strip ratio.

BTW, anybody care to chime in on what format for practice their club uses?

We use something similar (and we're also sabre). Basic format is training bouts: total of 10 touches, no score kept. Allows people to practice actions, without the pressure of competing against their teammates or feeling like they "have" to win. And, it allows for plenty of strip time, and a good rotation around. Its a good way to keep a group of people going on and on. Then, we'll throw in some sharp bouts towards the end of the evening.

For sharp bouts, we'll do those in a couple of different ways, to work on different tactical problems (such as starting out 8-3, or 8-7, or 10-5). Another thing we've been doing lately is a modified "King of the Hill" format with 5 touch bouts. If you manage to stay up for more than two bouts, your opponent will start out with a 1-0 advantage, then the next time, a 2-0 advantage, and so on, to a maximum of 4-0.

The idea is to keep it simple.

[brutus]

yes that would work and I know alot of fencers who do this, as do I sometimes. However better fencers do not always beat the inferior fencer, fencing a lesser is a different ball game to fencing someone better or the same level. Therefore 'cos you will come against lesser fencers in competition it makes sense to fence them regularly at club.

[the ancient one]

OK, so it was boring at work yesterday. I had offered one format for the proposed problem:

In one session:
2 pools of 3 fencers

pool 1 ABC 3 bouts
pool 2 DEU 3 bouts

In the next session the best fencer of pool 2 and worst fencer of pool 1 exchange positions.

So I worked the full scenario including the 3rd point above.

My proposed format seemed intuitively to improve the overall result. In the attached model you can see that if you do the exchange, only the D, the top fencer in the bottom pool benefits, and only a little. 4 others have a significantly decreased result.

The other problem, of course, is that we have not really tried to estimate how much experience it takes to get from one level to another. When I set the peak experience to 0.05 that would mean 20 identical optimum bouts=1 point which for certain is too low. I wish I could fence 20 bouts and increase 1 level! Any opinions for a good typical-average number? And a better starting level than 1,2,3,4,5,6?

It would be interesting to compare total career bouts at time of rating change with a little insight into how one learns to fence, classes,years under master, big club, small club etc... In the future, with things like the FIE database (if we had access), and even Ask Fred it could be possible to describe a "typical life-cycle".

[Falcon (XB :)]

To comment on Monash Armourer's post about the handicap system in place at that "large fencing club", bear in mind that that is put in place for one special Handicap Competition held in a year - for fun, not training, where the sole idea is to try and make it possible for children, beginner and novice fencers to have a reasonable chance against even the highest ranked national and international level fencers. Besides there is a heap of donated prizes and encouragement awards and the job of the handicapper is to ensure that it is unlikely that the best fencers in the country win the best prizes, but that a deserving, hard working youngster will.

Now, another handicapping system that can work if there are only a few fencers in at a training night and where there is quite a disparity in ability is to do a variation on the cumulative handicapping scheme.

If I am setting structured bouting scenarios for my junior students, I may have them fence 1-3-5-3-1 with cumulative handicaps. Bout 1 is to 1 point. Whoever wins starts at -1 in Bout 2 which is to 3 points. Obviously, if there is a big disparity in ability then one fencer may end up starting the Bout 5 (to 1 point) from -4. Great! Good for her, good for her opponent. If the non-handicapped fencer wins then they start the next bout at -1. Etc.

"But they're not training to 5 or 15 points. How unrealistic!" Oh, fiddle! What nonsense. You tell me what you are fencing to in a team match when you are 13-7 down in the 5th relay. Some coaches might say you have 3 minutes to try and score +6 if you can with an upper limit of 25. I.e. you can possibly score no more than 18 points but your aim is simply to cut the gap down.

Similarly within a 15 hit bout, you emerge after the 2nd break 12-8 up against a better fencer. What do you do. Have you practised this situation?

You might have if you had done the handy-dandy Cumulative Handicap 1-3-5-7-9-11-13-15-13-11-9-7-5-3-1 some rainy afternoon. Or the 1-3-5-3-1-3-5-3-1-3-5-3-1-3...[choose your end point] I leave it to Peter G to do the math as to the possible maximum or minimum number of contested points that must be scored in any of these scenarios. You might be surprised even in the smaller series. The breaks between bouts enable some recovery. The extra demand on the better fencer to score more points balances the difference in ability. There is still a psychological reward available for the weaker fencer to fend off the stronger fencer. Especially if you add a time factor.

E.g. The 2 Days in One World Cup Workout:

6 bouts of 5. Cumulative Handicap of -1 if a fencer wins a bout, to a max of -5 in Bout 6.

6 bouts of 15 (assume we're through to the 64 directly). CH of -3 per bout won.

Do it with a ref and a stopwatch. Or have a minute break at 8 if you are fencing foil or épée and you don't have a ref.

Now, here, we get the mental toughness factor, the fitness factor, the balance between disparate abilities, and, dare I say it...the fun factor. Works very well in preparing young athletes and when they can fence their coach for a block of time, for example. And excellent where there are few fencers available to fence and the ability levels are very diverse.

There are a heap more fun variations which serve similar purposes (such as the Tennis Set variations or the King of the Hill variations (I prefer with the -ve handicaps rather than the other way for that) aka Corwin on Kolvir when I'm setting the exercise [some of you will appreciate the reference]).

I very much appreciate Peter G's depth of thought about these issues. I am pleased to read every such post. Continue!

>^_^<